Changed EBI DbFetch code so that the function works again. Added latest picard jar. More work on CRAM file loading. Updated/added scripts for application start. Small tweaks to makefile/ant. Removed gff2embl and gene_builder. Recreated evosuite tests.

Initial BAM View changes to fix selection issues, zooming into secondary alignments, erroneous collapsing of reads with same start/end position and display of SNPs - related to RT ticket 597652

First commit of unit test changes. Added more manual unit tests, an EvoSuite regression set and coverage reporting. Moved Java version for Artemis to Java 8. Ripped out redundant Corba code.

update scratch108 to scratch118 - v16.0.18

Updated NCBI blast URL to https

version bump

add TCDB as database

fix method name clash for List.sort() in Java 8

override abstract method close() in CRAMReferenceSequenceFile

switch to older EMBOSS version

Building from source without this patch resulted in the following error:
artemis-16.0.0+dfsg/uk/ac/sanger/artemis/components/alignment/CRAMReferenceSequenceFile.java:35:
error: CRAMReferenceSequenceFile is not abstract and does not override abstract method close() in ReferenceSequenceFile

add new 'alias' qualifier

Added Kim Rutherford

Add TravisCI Support, change README to markdown

Rather than testing the Ant build under multiple JDKs, I've arbitrarily
decided to test it only under openjdk7.  This keeps us to 4 builds
rather than 6 before the build is reported as green which should be a
bit quicker.

ant builds copy everything in the pwd into ant_build and then try to
build all of the java files copied, including those of the dependencies
which are built separately.  This does not end well.